UCL Research Domains


Missing links: on the social contagion of child disappearances

2019-20 Social Science plus Pilot Project (£10,100)

Research question
The proposed project is a collaboration between UCL, Thames Valley Police and the charity Missing People. It grew out of three converging factors: (1) increases in the number of children going missing; (2) advancements in social network analysis (SNA), and (3) the significant insights gleaned from applying SNA to understand patterns of exploitation and victimisation.

Through privileged access to a large multiagency dataset, this study will determine the patterns and predictors of missing incidents involving children, with a focus on examining empirically whether a child’s social environment influences their likelihood of going missing. In particular, it will examine whether the risk of going missing is transmissible through social networks, in such a way that missing episodes among individuals might be predictive of subsequent episodes amongst their acquaintances.

Focus, rationale and societal relevance
One person goes missing every two minutes in England and Wales, and responding to reports of missing persons is a major source of police demand, estimated to cost upward of £500 million annually1. Children account for the majority of missing incidents, with many going missing repeatedly2. The police place a high priority on the investigation of missing children in light of emerging evidence that going missing is apotential indicator of abuse, sexual and criminal exploitation and harm3.

There are various ‘push’ and ‘pull’ factors that account for why children go missing4. One common hypothesis is that the likelihood of a child going missing is influenced by their social environment, in particular their exposure to individuals who have previously gone missing. This echoes research from the UK which shows that young people’s peer relations can, for example, spread and sustain child sexual exploitation5.
Recent research in the US has demonstrated that the risk of violent crime shows evidence of transmissibility through social networks6. Specifically, it has shown that an individual is more likely to be the victim of gun violence if one of their social contacts has recently been victimised, suggesting that risk is communicable7. This has clear implications for the identification and management of risk, and invites a ‘public health’ approach to harm prevention. These ideas have not yet, however, been applied to other forms of harm and victimisation, including the problem of missing children.

Research design and methodology
The proposed research has three stages:
Phase 1 – data cleaning and preparation
: The research will use a dataset provided by Thames Valley Police containing information on approximately 120,000 individuals living in Reading, UK. As well as data concerning events with police involvement (including missing episodes), the dataset incorporates information relating to education, social care and housing. These datasets will be used to derive social networks reflecting a range of relationships: as well as familial links, individuals will be connected if they share the same school class, for example, or a common care home.

Phase 2 – data analysis and hypothesis testing: Descriptive analysis will establish the extent and nature of peer networks of missing children and examine clustering within it. Multilevel logistic regression will then be carried out to determine the probability of a child going missing (repeatedly) as a function of network influences (exposure/proximity to other missing persons) and individual-level variables (age, sex, housing status). Further modelling (via Hawkes process) will then be used to measure the extent and duration of any observed contagion effects.

Phase 3 – interpretation and translation: As this is a collaborative project with police and charity partners, the final stage will be dedicated to maximising the impact of our findings for police and partner agencies’ affected by the problem of missing persons. In particular, we will hold a one-day workshop in London, which will also be used to guide the development of, and build support for, our subsequent funding bid.

Research Team
Social Scientist Principle Investigator
Dr. Aiden Sidebottom, Senior Lecturer, Security & Crime Science, Engineering Sciences, BEAMS

Non-Social Science Co-Investigator
Dr. Toby Davies, Lecturer, Security & Crime Science, Engineering Sciences, BEAMS

Additional Collaborators
Thames Valley Police
Missing People

1 Babuta, A., & Sidebottom, A. (2019). Missing Children: On the Extent, Patterns, and Correlates of Repeat Disappearances by Young People. Policing: A Journal of Policy and Practice.

2 Sidebottom, A., Boulton, L., Cockbain, E., Halford, E., & Phoenix, J. (2019). Missing children: risks, repeats and responses. Policing and Society, 1-14.

3 Sharp-Jeffs, N., 2016. Hidden links? Going missing as an indicator of child sexual exploitation. In: K. Shalev-Greene and L. Alys, eds. Missing persons: a handbook of research. New York, NY: Taylor &Francis, 18–27.

4 Biehal, N., Mitchell, F., & Wade, J. (2003). Lost from view: Missing persons in the UK. Policy Press.

5 Cockbain, E. (2018). Offender and victim networks in human trafficking. Routledge.

6 Papachristos, A. V., Braga, A. A., & Hureau, D. M. (2012). Social networks and the risk of gunshot injury. Journal of Urban Health, 89(6), 992-1003.

7 Green, B., Horel, T., & Papachristos, A. V. (2017). Modeling contagion through social networks to explain and predict gunshot violence in Chicago, 2006 to 2014. JAMA internal medicine, 177(3), 326-333.